geointel/scripts/provision_flanders_geographic_scope.py

"""Provision all current official Flemish municipality Areas.

The municipality inventory is discovered from the official VRBG RefGem
collection at runtime. The script uses the existing geographic-scope artifact
and API persistence flow; it never writes directly to PostGIS.
"""

from __future__ import annotations

import argparse
import json
import os
import sys
from datetime import datetime, timezone
from pathlib import Path
from typing import Any

import requests

from geographic_scopes import GeographicScope, ScopeMember
from provision_geographic_scope import (
    VRBG_ATTRIBUTION,
    VRBG_ITEMS_URL,
    build_scope_payloads,
    build_source_session,
    provision_scope,
    sha256_file,
    write_json_atomic,
)


DEFAULT_API_URL = "http://127.0.0.1:8000"
DEFAULT_OUTPUT_ROOT = Path("/app/storage/operator-data/geographic-scopes")
MIN_EXPECTED_MUNICIPALITIES = 270
MAX_EXPECTED_MUNICIPALITIES = 300


def parse_args() -> argparse.Namespace:
    parser = argparse.ArgumentParser(description="Provision the current official Flanders municipality scope.")
    parser.add_argument("--base-url", default=os.environ.get("GEOINTEL_INTERNAL_API_URL", DEFAULT_API_URL))
    parser.add_argument(
        "--output-root",
        type=Path,
        default=Path(os.environ.get("GEOINTEL_SCOPE_OUTPUT_ROOT", DEFAULT_OUTPUT_ROOT)),
    )
    parser.add_argument("--request-timeout", type=int, default=180)
    parser.add_argument("--import-timeout", type=int, default=3600)
    parser.add_argument("--force", action="store_true")
    parser.add_argument("--fetch-only", action="store_true")
    parser.add_argument("--min-municipalities", type=int, default=MIN_EXPECTED_MUNICIPALITIES)
    parser.add_argument("--max-municipalities", type=int, default=MAX_EXPECTED_MUNICIPALITIES)
    return parser.parse_args()


def discover_flanders_scope(
    session: requests.Session,
    *,
    timeout: int,
    min_municipalities: int,
    max_municipalities: int,
) -> tuple[GeographicScope, list[dict[str, Any]], str]:
    if min_municipalities <= 0 or max_municipalities < min_municipalities:
        raise ValueError("Municipality count safety limits are invalid")
    response = session.get(
        VRBG_ITEMS_URL,
        params={"f": "application/geo+json", "limit": "1000"},
        timeout=timeout,
    )
    response.raise_for_status()
    payload = response.json()
    features = payload.get("features")
    if not isinstance(features, list):
        raise RuntimeError("Official VRBG response does not contain a feature list")
    if not min_municipalities <= len(features) <= max_municipalities:
        raise RuntimeError(
            f"Official VRBG returned {len(features)} municipalities; expected "
            f"{min_municipalities}..{max_municipalities}. Refusing an incomplete or broadened scope."
        )

    by_code: dict[str, dict[str, Any]] = {}
    names: set[str] = set()
    for feature in features:
        if not isinstance(feature, dict) or not isinstance(feature.get("geometry"), dict):
            raise RuntimeError("Official VRBG contains a malformed municipality feature")
        properties = feature.get("properties")
        if not isinstance(properties, dict):
            raise RuntimeError("Official VRBG municipality is missing properties")
        nis_code = str(properties.get("NISCODE") or "").strip()
        name = str(properties.get("NAAM") or "").strip()
        if len(nis_code) != 5 or not nis_code.isdigit() or not name:
            raise RuntimeError(f"Official VRBG municipality identity is invalid: {nis_code!r} / {name!r}")
        if nis_code in by_code or name.casefold() in names:
            raise RuntimeError(f"Official VRBG contains a duplicate municipality: {nis_code} / {name}")
        by_code[nis_code] = feature
        names.add(name.casefold())

    ordered = [by_code[code] for code in sorted(by_code)]
    members = tuple(
        ScopeMember(
            name=str(feature["properties"]["NAAM"]).strip(),
            nis_code=str(feature["properties"]["NISCODE"]).strip(),
        )
        for feature in ordered
    )
    scope = GeographicScope(
        key="flanders",
        display_name=f"Vlaanderen ({len(members)} gemeenten)",
        project_name="Flanders Regional Workbench",
        project_region="Vlaanderen, België",
        area_name="Vlaanderen - officiële operationele grens",
        authority_name="Digitaal Vlaanderen VRBG",
        authority_url=VRBG_ITEMS_URL,
        scope_type="region",
        limitation_message=(
            "Officiële unie van de actuele VRBG-gemeentegrenzen. De operationele regio omvat het Vlaamse "
            "landgebied; territoriale zee, EEZ en continentaal plat zijn afzonderlijke maritieme scopes."
        ),
        members=members,
    )
    return scope, ordered, response.url


def prepare_artifacts(
    args: argparse.Namespace,
    scope: GeographicScope,
    features: list[dict[str, Any]],
    source_url: str,
) -> tuple[Path, Path, dict[str, Any]]:
    output_dir = args.output_root / scope.key
    manifest_path = output_dir / "flanders_scope_manifest.json"
    generated_at = datetime.now(timezone.utc).isoformat()
    snapshot_date = generated_at[:10]
    boundary_path = output_dir / f"flanders_boundary_{snapshot_date}.geojson"
    members_path = output_dir / f"flanders_municipalities_{snapshot_date}.geojson"
    member_codes = list(scope.nis_codes)

    if not args.force and manifest_path.is_file():
        manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
        cached_boundary = output_dir / str(manifest.get("boundary_filename") or "")
        cached_members = output_dir / str(manifest.get("municipalities_filename") or "")
        if (
            manifest.get("status") == "complete"
            and manifest.get("member_nis_codes") == member_codes
            and cached_boundary.is_file()
            and cached_members.is_file()
            and sha256_file(cached_boundary) == manifest.get("boundary_sha256")
            and sha256_file(cached_members) == manifest.get("municipalities_sha256")
        ):
            return cached_boundary, cached_members, manifest

    boundary_payload, members_payload, summary = build_scope_payloads(
        scope,
        features,
        source_url=source_url,
        generated_at=generated_at,
    )
    write_json_atomic(boundary_path, boundary_payload)
    write_json_atomic(members_path, members_payload)
    manifest = {
        "schema_version": 1,
        "status": "complete",
        "generated_at": generated_at,
        "observed_at": f"{snapshot_date}T00:00:00Z",
        "boundary_filename": boundary_path.name,
        "boundary_sha256": sha256_file(boundary_path),
        "municipalities_filename": members_path.name,
        "municipalities_sha256": sha256_file(members_path),
        "vrbg_source_url": source_url,
        "vrbg_attribution": VRBG_ATTRIBUTION,
        "scope_authority_name": scope.authority_name,
        "scope_authority_url": scope.authority_url,
        "scope_limitation": scope.limitation_message,
        **summary,
    }
    write_json_atomic(manifest_path, manifest, pretty=True)
    return boundary_path, members_path, manifest


def main() -> int:
    args = parse_args()
    args.operator_tool = "provision_flanders_geographic_scope.py"
    try:
        with build_source_session() as session:
            scope, features, source_url = discover_flanders_scope(
                session,
                timeout=args.request_timeout,
                min_municipalities=args.min_municipalities,
                max_municipalities=args.max_municipalities,
            )
        boundary_path, members_path, manifest = prepare_artifacts(args, scope, features, source_url)
        workspace = None if args.fetch_only else provision_scope(
            args, scope, boundary_path, members_path, manifest
        )
    except (OSError, RuntimeError, ValueError, KeyError, requests.RequestException) as exc:
        print(json.dumps({"status": "error", "message": str(exc)}, ensure_ascii=False), file=sys.stderr)
        return 1

    print(
        json.dumps(
            {
                "status": "ok",
                "mode": "fetch_only" if args.fetch_only else "provisioned",
                "scope": scope.key,
                "member_count": manifest["member_count"],
                "area_km2": manifest["area_km2"],
                "wgs84_bbox": manifest["wgs84_bbox"],
                "boundary_path": str(boundary_path),
                "municipalities_path": str(members_path),
                "workspace": workspace,
                "limitation": scope.limitation_message,
            },
            ensure_ascii=False,
            indent=2,
        )
    )
    return 0


if __name__ == "__main__":
    raise SystemExit(main())